470 



SCIENCE. 



fX. S. Vol. XVIII. Xo. 458. 



secondary schools has but few ardent sup- 

 porters. 



In teaching chemistry in graded and sec- 

 ondary schools would it not be more profit- 

 able to spend the time in the laboratory 

 on pure descriptive chemistry, and could 

 not the time spent on abstract chemical 

 theory more profitably be spent emphasiz- 

 ing the relations expressed in the phrase 

 'The Chemistry of Daily Life'? 



Francis Gano Benedict. 



aOIENTIFW BOOKS. 

 Chemical Analyses of Igneous Rocks Pub- 

 lished from 188Jf to 1900, with a critical 

 discussion of the character and use of anal- 

 yses. By Henry Stephens Washington. 

 United States Geological Survey, Profes- 

 sional Paper No. 14. Washington, Govern- 

 ment Printing Office. 1903. Pp. 495. 

 In the first two or three decades of the last 

 century, rocks, in contradistinction to the 

 individually well-defined minerals, were re- 

 garded merely as aggregates of minerals in 

 presumably fortuitous combinations and 

 lacking in that definiteness or constancy of 

 composition which would justify their study 

 as a whole. As time went on, however, the 

 chemical aspect of petrography gi-advially at- 

 tracted more attention, its great importance 

 being first clearly recognized by Abrich, who 

 in 1841 pointed out the necessity for a 

 knowledge of the chemical composition of 

 rocks in dealing with the problems of their 

 origin and mutual relations as well as for 

 their satisfactory classification and proper 

 nomenclature. 



Bunsen's well-known hypothesis that all 

 igneous rocks might be considered as mix- 

 tures in various proportions of two supposed 

 original or normal magmas — the trachytic 

 and the pyroxenic — gave a marked stimulus 

 to the study of their chemical composition. 

 But with the abandonment of this view, which 

 broke down under the weight of the evidence 

 accumulated to test it and with the introduc- 

 tion of the microscope as a means of petro- 

 graphical study in the early seventies, anal- 



yses lost much of their interest, being, as 

 Dr. Washington observes, ' inserted perfunc- 

 torily in petrographical writings, in obedience 

 to custom, as ornamental embellishments, 

 while the chief efforts of the petrographer 

 were devoted to the elucidation of the purely 

 mineralogical and textural characters of the 

 rocks described.' 



During the past ten years, however, the 

 chemical composition of rocks has again at- 

 tracted much attention — more especially on 

 account of its important bearing on the 

 theoretical side of petrography — the crystal- 

 lographic and optical properties of the con- 

 stituent minerals and the details of structure 

 no longer being the only subject-s of investi- 

 gation. The chemical composition of an 

 igneous rock is now recognized as distinctly 

 the most important fact which can be learned 

 concerning it, and the one which is of the 

 greatest value in dealing with the great ques- 

 tions of origin and genetic relations, as well 

 as affording the most reliable basis for classi- 

 fication. 



Por the study of the chemical composition 

 of rocks, the tables of analyses collected by 

 Eoth and which were issued at intervals from 

 1861 to 1884, have up to the present been the 

 great storehouse of information. They pre- 

 sent in tabular form and with certain critical 

 notes practically all the rock analyses which 

 had been published up to the year 1884. Since 

 this latter year, however, a great number of 

 analyses have appeared, in widely scatteretl 

 journals, proceedings and reports, showing a 

 marked improvement in quality as compared 

 with the old analyses. 



In the present volume Dr. Washington has 

 collected all the analyses which have been pub- 

 lished during the seventeen years from 1883 

 to 1900, and has presented them excellently 

 arranged according to the quantitative system 

 of classification recently proposed by him in 

 conjunction with Iddings, Pirsson and Cross, 

 together with full references and with critical 

 notes when required, the whole being intro- 

 duced by an admirable series of chapters deal- 

 ing with the character of rock analyses and 

 their bearing on rock classification. The vol- 



